AbstractThe purpose of this study is to investigate the validity of performing biochemical analyses on human endomyocardlal biopsies. One problem of such analyses is the decay of metabolites between the time of biopsy and the analyses. This was addressed by varying the time delay from biopsy to submersion in liquid nitrogen (‐200°C). Neither cyclic AMP or cyclic GMP changed significantly up to 90 seconds after biopsy. The reproduclblllty of the assays in human heart biopsies was determined by submitting paired samples from each procedure. The ATP content was determined by a fluorescent technique and was related to the protein and creative content. The average ATP of 25 paired samples was 12.3 nm per mg protein; the average difference between the paired samples was ‐0.6, and the 95%confidence Interval was 5.6 nm/mg protein. The average ATP‐to‐creatJne ratio was 0.25; the 95% confidence interval was 0.12. The cyclic AMP and cyclic GMP were found to average 3.3 and 0.25 femtomoles/mcg protein, respectively. The 95% confidence intervals were 1.8 and 0.12 fmoles/mcg protein. The problems of sampling error and blood and connective tissue contamination decrease the reliability of the biochemical analyses. Reproduclblllty is adequate, however, to acquire meaningful biochemical information on the hum