Anemic conditions in women and children across the world are a serious cause for concern. The standard method for measuring hemoglobin (Hb) in human blood is the well-recognized cyanmethemoglobin method as recommended by the World Health Organization (WHO). There are a number of methods available that give very approximate results as compared to this method. The color-measurement technique is one of them and it is recommended by the WHO for adaptation in low-resource settings. Since human interpretation errors are likely to creep in during the subjective processes involved with this method, an artificial neural network (ANN) approach for the estimation of Hb count in human blood has been evaluated. The ANN used color-coded values of the samples as input and the Hb value, as obtained with the cyanmethemoglobin method, as desired output using 2007 samples. The results show a strong relation between the color of the blood sample and the Hb level in the blood.