This work aimed to estimate the growth curve, by means of non-linear models, of 399 Boer X Anglonubian crossbred goats bred in the microregion of Teresina, Piauí State, Brasil, born between 2006 and 2010, and environmental and genetic factors affecting this curve. The adjusted models were Gompertz, Logistic, Von Bertalanffy, Brody and Richards. After adjustment of these models, a study of the environmental effects on the growth curve was proceeded, considering of year, season, type of birth, sex, and herd effects. The averages of weights were low indicating reflection of the adopted range breeding system. The Gompertz model presented the best fit according to the criteria QMR, R², percentage of convergence, behavior chart and DMA. Significant influence was found (p<0,05) of year and season of birth, type of birth, sex, and flock effects on the A, B and k parameters of Growth curve adjusted and on the estimated weights from the fitting curve. The Absolute Maximum Growth Rate was early observed at 35 days old, indicating that the animals suffered nutritional deficits which reflect on their growth, especially in the post-weaning. The estimated heritabilities for curve parameters were of low magnitude, resulting in little change in the growth pattern of the animals studied by means of individual selection. The low genetic variance for the parameters of the curve it is because of the limits of thesample database.Objetivou-se com este trabalho estimar a curva de crescimento, por meio de modelos não-lineares, de 399 caprinos mestiços Boer x Anglonubiana criados extensivamente na microrregião de Teresina, PI, nascidos entre 2006 e 2010, e estudar fatores ambientais e genéticos que influenciam a curva de crescimento. Os modelos ajustados foram Gompertz, Logístico, Von Bertalanffy, Brody e Richards. Após ajuste dos modelos procedeu-se estudo dos efeitos ambientais sobre a curva de crescimento, sendo considerados os efeitos de ano, estação, tipo de nascimento, sexo e rebanho. As médias dos pesos foram baixas, indicando reflexo do sistema de criação extensivo adotado. O modelo de Gompertz apresentou melhor ajuste, de acordo com os critérios QMR, R², percentual de convergência, comportamento gráfico e DMA. Constatou-se influência significativa (p<0,05) dos efeitos de ano e estação de nascimento, tipo de nascimento, sexo e rebanho sobre os parâmetros A, B e k da curva de crescimento ajustada e sobre os pesos estimados a partir da curva ajustada. A Taxa de Crescimento Absoluta máxima foi observada precocemente aos 35 dias de idade, indicando que os animais sofreram déficit nutricional que refletem no crescimento, principalmente na fase pós desmama. As herdabilidades estimadas para os parâmetros da curva foram de baixa magnitude, resultando em pequena alteração do padrão de crescimento dos animais estudados por meio de seleção individual. A baixa variância genética para os parâmetros da curva ajustada é conseqüência das limitações amostrais do banco de dados. Environmental and genetic factors on the growth curve of crossbred goatsThis work aimed to estimate the growth curve, by means of non-linear models, of 399 Boer XAnglonubian crossbred goats bred in the microregion of Teresina, Piauí State, Brasil, born between2006 and 2010, and environmental and genetic factors affecting this curve. The adjusted modelswere Gompertz, Logistic, Von Bertalanffy, Brody and Richards. After adjustment of these models,a study of the environmental effects on the growth curve was proceeded, considering of year,season, type of birth, sex, and herd effects. The averages of weights were low indicating reflectionof the adopted range breeding system. The Gompertz model presented the best fit according tothe criteria QMR, R², percentage of convergence, behavior chart and DMA. Significant influencewas found (p<0,05) of year and season of birth, type of birth, sex, and flock effects on the A, B andk parameters of Growth curve adjusted and on the estimated weights from the fitting curve. TheAbsolute Maximum Growth Rate was early observed at 35 days old, indicating that the animalssuffered nutritional deficits which reflect on their growth, especially in the post-weaning. Theestimated heritabilities for curve parameters were of low magnitude, resulting in little change in thegrowth pattern of the animals studied by means of individual selection. The low genetic variance forthe parameters of the curve it is because of the limits of thesample database.