The possibility of false-negative results in clenbuterol analysis was investigated in bovine tissues. An extraction procedure currently in use was adapted to process 100 specimens of different tissues each time. Its efficiency and accuracy were investigated radiometrically by means of a series of different molar concentration of the tritiated drug. In samples not submitted to extensive delipidation, unreliability of the analysis was evident. The measurement of tissue clenbuterol content, by a competitive ELISA, gave results numerically similar to those existing in literature, but with an accuracy high enough to minimize the frequency of false-negative results.