This paper compares the performance of the program evaluation phase of genetic programming using C and Common Lisp. A simple experiment is conducted, and the conclusion is that genetic programming implemented in Common Lisp using on-the-fly compilation of the evolved programs can be faster than an implementation in C, also when the compilation time is taken into consideration. The deciding factor is the number of times that each evolved program is evaluated. Categories and Subject Descriptors I.2.6 [Artificial Intelligence]: Learning. General Terms Performance, Experimentation, Languages. Keywords Genetic Programming, Lisp, C, Implementation, Performance