RESUMOObjetivos: O objetivo deste estudo é analisar a validade das equações de predição da taxa metabólica de repouso (TMR) propostas por Harris e Benedicti (1919), Schofield (1985), WHO/FAO/UNU (1985), Henry e Rees (1991), Molnár e cols. (1995), Tverskaya e cols. (1998) e Müller e cols. (2004) em adolescentes meninas. Sujeitos e métodos: Para tanto, avaliaram-se 51 meninas entre 10 e 17 anos, sendo mensuradas as seguintes variáveis: massa corporal, estatura, massa gorda (absoluta e percentual) e massa magra (absoluta e percentual), bem como os valores do consumo de VO 2 e produção de VCO 2 que foram substituídos na equação de Weir (1949) considerada, neste estudo, como método padrão da TMR. Realizaram-se os seguintes procedimentos estatísticos: teste "t" pareado (p < 0,05); erro constante (EC) com diferença menor que 5% e análise gráfica de Bland e Altman (1986). Resultados: As equações propostas por Tverskaya e cols. (1998) The following anthropometric variables were measured: body weight, stature, body fat mass (absolute and percentage), body lean mass (absolute and percentage) as well as the consumption of VO 2 and production of VCO 2 that were substituted in Weir's (1949) equation which was considered the RMR standard method in this study. The following statistical procedures were carried out: paired t-test (p < 0.05); constant error (CE) with a less than 5% difference and Bland and Altman's (1986) graphical analysis. Results: The equations proposed by Tverskaya and cols. (1998) and Müller and cols. (2004) were the only equations that presented significant differences between the averages. However, only the equations proposed by Henry and Rees (1991) and Molnár and cols. (1995) produced a constant error of less than 5%. However, Bland and Altman's (1986) graphical analysis showed that independently of the equation there is no agreement between the equations and the standard method. Conclusions: The study indicates that none of the equations meet all the stipulated criteria and that thus have no validity in the calculation of the RMR. Arq Bras Endocrinol Metab. 2010;54(1):30-6