fpzip is a library and command-line utility for lossless and optionally lossy compression of 2D and 3D floating-point arrays. fpzip assumes spatially correlated scalar-valued data, such as regularly sampled continuous functions, and is not suitable for compressing unstructured streams of floating-point numbers. In lossy mode, fpzip discards some number of least significant mantissa bits and losslessly compresses the result. fpzip currently supports IEEE-754 single (32-bit) and double (64-bit) precision floating-point data. fpzip is written in C++ but has a C compatible API that can be called from C and other languages. It conforms to the C++98 and C89 language standards.