diff --git a/docs/src/api_bindings.md b/docs/src/api_bindings.md
index 98b882056..3a43a129c 100644
--- a/docs/src/api_bindings.md
+++ b/docs/src/api_bindings.md
@@ -87,6 +87,7 @@ h5a_write
 ## [[`H5D`](https://portal.hdfgroup.org/display/HDF5/Datasets) — Dataset Interface](@id H5D)
 - [`h5d_close`](@ref h5d_close)
 - [`h5d_create`](@ref h5d_create)
+- [`h5d_create_anon`](@ref h5d_create_anon)
 - [`h5d_extend`](@ref h5d_extend)
 - [`h5d_fill`](@ref h5d_fill)
 - [`h5d_flush`](@ref h5d_flush)
@@ -116,6 +117,7 @@ h5a_write
 ```@docs
 h5d_close
 h5d_create
+h5d_create_anon
 h5d_extend
 h5d_fill
 h5d_flush
diff --git a/gen/api_defs.jl b/gen/api_defs.jl
index ed06371ba..e1f1d15c0 100644
--- a/gen/api_defs.jl
+++ b/gen/api_defs.jl
@@ -56,6 +56,7 @@
 
 @bind h5d_close(dataset_id::hid_t)::herr_t "Error closing dataset"
 @bind h5d_create2(loc_id::hid_t, pathname::Ptr{UInt8}, dtype_id::hid_t, space_id::hid_t, lcpl_id::hid_t, dcpl_id::hid_t, dapl_id::hid_t)::hid_t string("Error creating dataset ", h5i_get_name(loc_id), "/", pathname)
+@bind h5d_create_anon(loc_id::hid_t, type_id::hid_t, space_id::hid_t, dcpl_id::hid_t, dapl_id::hid_t)::hid_t "Error in creating anonymous dataset"
 @bind h5d_extend(dataset_id::hid_t, size::Ptr{hsize_t})::herr_t "Error extending dataset" # deprecated in favor of h5d_set_extent
 @bind h5d_fill(fill::Ptr{Cvoid}, fill_type_id::hid_t, buf::Ptr{Cvoid}, buf_type_id::hid_t, space_id::hid_t)::herr_t "Error filling dataset"
 @bind h5d_flush(dataset_id::hid_t)::herr_t "Error flushing dataset"
diff --git a/src/HDF5.jl b/src/HDF5.jl
index db46f1a7f..769b037c8 100644
--- a/src/HDF5.jl
+++ b/src/HDF5.jl
@@ -567,7 +567,8 @@ end
 
 # Arguments
 * `parent` - `File` or `Group`
-* `path` - String describing the path of the dataset within the HDF5 file
+* `path` - `String` describing the path of the dataset within the HDF5 file or
+           `nothing` to create an anonymous dataset
 * `datatype` - `Datatype` or `Type` or the dataset
 * `dataspace` - `Dataspace` or `Dims` of the dataset
 * `properties` - keyword name-value pairs set properties of the dataset
@@ -589,7 +590,7 @@ See also
 """
 function create_dataset(
     parent::Union{File,Group},
-    path::AbstractString,
+    path::Union{AbstractString,Nothing},
     dtype::Datatype,
     dspace::Dataspace;
     dcpl::DatasetCreateProperties = DatasetCreateProperties(),
@@ -597,14 +598,19 @@ function create_dataset(
     dapl::DatasetAccessProperties = DatasetAccessProperties(),
     pv...
 )
-    haskey(parent, path) && error("cannot create dataset: object \"", path, "\" already exists at ", name(parent))
+    !isnothing(path) && haskey(parent, path) && error("cannot create dataset: object \"", path, "\" already exists at ", name(parent))
     pv = setproperties!(dcpl,dxpl,dapl; pv...)
     isempty(pv) || error("invalid keyword options")
-    Dataset(API.h5d_create(parent, path, dtype, dspace, _link_properties(path), dcpl, dapl), file(parent), dxpl)
+    if isnothing(path)
+        ds = API.h5d_create_anon(parent, dtype, dspace, dcpl, dapl)
+    else
+        ds = API.h5d_create(parent, path, dtype, dspace, _link_properties(path), dcpl, dapl)
+    end
+    Dataset(ds, file(parent), dxpl)
 end
-create_dataset(parent::Union{File,Group}, path::AbstractString, dtype::Datatype, dspace_dims::Dims; pv...) = create_dataset(checkvalid(parent), path, dtype, dataspace(dspace_dims); pv...)
-create_dataset(parent::Union{File,Group}, path::AbstractString, dtype::Datatype, dspace_dims::Tuple{Dims,Dims}; pv...) = create_dataset(checkvalid(parent), path, dtype, dataspace(dspace_dims[1], max_dims=dspace_dims[2]); pv...)
-create_dataset(parent::Union{File,Group}, path::AbstractString, dtype::Type, dspace_dims::Tuple{Dims,Dims}; pv...) = create_dataset(checkvalid(parent), path, datatype(dtype), dataspace(dspace_dims[1], max_dims=dspace_dims[2]); pv...)
+create_dataset(parent::Union{File,Group}, path::Union{AbstractString,Nothing}, dtype::Datatype, dspace_dims::Dims; pv...) = create_dataset(checkvalid(parent), path, dtype, dataspace(dspace_dims); pv...)
+create_dataset(parent::Union{File,Group}, path::Union{AbstractString,Nothing}, dtype::Datatype, dspace_dims::Tuple{Dims,Dims}; pv...) = create_dataset(checkvalid(parent), path, dtype, dataspace(dspace_dims[1], max_dims=dspace_dims[2]); pv...)
+create_dataset(parent::Union{File,Group}, path::Union{AbstractString,Nothing}, dtype::Type, dspace_dims::Tuple{Dims,Dims}; pv...) = create_dataset(checkvalid(parent), path, datatype(dtype), dataspace(dspace_dims[1], max_dims=dspace_dims[2]); pv...)
 
 # Note that H5Tcreate is very different; H5Tcommit is the analog of these others
 create_datatype(class_id, sz) = Datatype(API.h5t_create(class_id, sz))
@@ -645,7 +651,7 @@ move_link(parent::Union{File,Group}, src_name::AbstractString, dest_name::Abstra
 Base.setindex!(dset::Dataset, val, name::AbstractString) = write_attribute(dset, name, val)
 Base.setindex!(x::Attributes, val, name::AbstractString) = write_attribute(x.parent, name, val)
 # Create a dataset with properties: obj[path, prop = val, ...] = val
-function Base.setindex!(parent::Union{File,Group}, val, path::AbstractString; pv...)
+function Base.setindex!(parent::Union{File,Group}, val, path::Union{AbstractString,Nothing}; pv...)
     need_chunks = any(k in keys(chunked_props) for k in keys(pv))
     have_chunks = any(k == :chunk for k in keys(pv))
 
@@ -1170,7 +1176,7 @@ function readmmap(obj::Dataset)
 end
 
 # Generic write
-function Base.write(parent::Union{File,Group}, name1::AbstractString, val1, name2::AbstractString, val2, nameval...) # FIXME: remove?
+function Base.write(parent::Union{File,Group}, name1::Union{AbstractString,Nothing}, val1, name2::Union{AbstractString,Nothing}, val2, nameval...) # FIXME: remove?
     if !iseven(length(nameval))
         error("name, value arguments must come in pairs")
     end
@@ -1191,7 +1197,7 @@ end
 # Create datasets and attributes with "native" types, but don't write the data.
 # The return syntax is: dset, dtype = create_dataset(parent, name, data; properties...)
 
-function create_dataset(parent::Union{File,Group}, name::AbstractString, data; pv...)
+function create_dataset(parent::Union{File,Group}, name::Union{AbstractString,Nothing}, data; pv...)
     dtype = datatype(data)
     dspace = dataspace(data)
     obj = try
@@ -1213,7 +1219,7 @@ function create_attribute(parent::Union{File,Object}, name::AbstractString, data
 end
 
 # Create and write, closing the objects upon exit
-function write_dataset(parent::Union{File,Group}, name::AbstractString, data; pv...)
+function write_dataset(parent::Union{File,Group}, name::Union{AbstractString,Nothing}, data; pv...)
     obj, dtype = create_dataset(parent, name, data; pv...)
     try
         write_dataset(obj, dtype, data)
@@ -1259,7 +1265,7 @@ function Base.write(obj::Dataset, x)
 end
 
 # For plain files and groups, let "write(obj, name, val; properties...)" mean "write_dataset"
-Base.write(parent::Union{File,Group}, name::AbstractString, data; pv...) = write_dataset(parent, name, data; pv...)
+Base.write(parent::Union{File,Group}, name::Union{AbstractString,Nothing}, data; pv...) = write_dataset(parent, name, data; pv...)
 # For datasets, "write(dset, name, val; properties...)" means "write_attribute"
 Base.write(parent::Dataset, name::AbstractString, data; pv...) = write_attribute(parent, name, data; pv...)
 
@@ -1562,7 +1568,7 @@ function write_dataset(dataset::Dataset, memtype::Datatype, buf::AbstractArray,
     stride(buf, 1) != 1 && throw(ArgumentError("Cannot write arrays with a different stride than `Array`"))
     API.h5d_write(dataset, memtype, API.H5S_ALL, API.H5S_ALL, xfer, buf)
 end
-function write_dataset(dataset::Dataset, memtype::Datatype, str::AbstractString, xfer::DatasetTransferProperties=dataset.xfer)
+function write_dataset(dataset::Dataset, memtype::Datatype, str::Union{AbstractString,Nothing}, xfer::DatasetTransferProperties=dataset.xfer)
     strbuf = Base.cconvert(Cstring, str)
     GC.@preserve strbuf begin
         # unsafe_convert(Cstring, strbuf) is responsible for enforcing the no-'\0' policy,
diff --git a/src/api/functions.jl b/src/api/functions.jl
index 55ea58dce..c76e3a763 100644
--- a/src/api/functions.jl
+++ b/src/api/functions.jl
@@ -307,6 +307,17 @@ function h5d_create(loc_id, pathname, dtype_id, space_id, lcpl_id, dcpl_id, dapl
     return var"#status#"
 end
 
+"""
+    h5d_create_anon(loc_id::hid_t, type_id::hid_t, space_id::hid_t, dcpl_id::hid_t, dapl_id::hid_t) -> hid_t
+
+See `libhdf5` documentation for [`H5Dcreate_anon`](https://portal.hdfgroup.org/display/HDF5/H5D_CREATE_ANON).
+"""
+function h5d_create_anon(loc_id, type_id, space_id, dcpl_id, dapl_id)
+    var"#status#" = ccall((:H5Dcreate_anon, libhdf5), hid_t, (hid_t, hid_t, hid_t, hid_t, hid_t), loc_id, type_id, space_id, dcpl_id, dapl_id)
+    var"#status#" < 0 && @h5error("Error in creating anonymous dataset")
+    return var"#status#"
+end
+
 """
     h5d_extend(dataset_id::hid_t, size::Ptr{hsize_t})
 
diff --git a/test/plain.jl b/test/plain.jl
index cc129ad89..3d5de9d3d 100644
--- a/test/plain.jl
+++ b/test/plain.jl
@@ -36,6 +36,8 @@ f["compressedempty", shuffle=true, deflate=4] = Int64[]
 # compression of zero-dimensional array (pull request #445)
 f["compressed_zerodim", shuffle=true, deflate=4] = fill(Int32(42), ())
 f["bloscempty", blosc=4] = Int64[]
+# test creation of an anonymouse dataset
+f[nothing] = 5
 # Create arrays of different types
 A = randn(3, 5)
 write(f, "Afloat64", convert(Matrix{Float64}, A))
@@ -86,6 +88,7 @@ write_attribute(f["int_vlen"], "vlen_attr", vleni)
 # Empty arrays
 empty = UInt32[]
 write(f, "empty", empty)
+write(f, nothing, empty)
 # Empty strings
 empty_string = ""
 write(f, "empty_string", empty_string)
@@ -133,6 +136,8 @@ Xslab = randn(20, 20, 5)
 for i = 1:5
     dset[:,:,i] = Xslab[:,:,i]
 end
+dset = create_dataset(f, nothing, datatype(Float64), dataspace(20, 20, 5), chunk=(5, 5, 1))
+dset[:, :, :] = 3.0
 # More complex hyperslab and assignment with "incorrect" types (issue #34)
 d = create_dataset(f, "slab2", datatype(Float64), ((10, 20), (100, 200)), chunk=(1, 1))
 d[:,:] = 5