Abstract The modeling of fertility patterns is an essential method researchers use to understand world-wide population patterns. Various types of fertility models have been reported in the literature to capture the patterns specific to developed countries. While much effort has been put into reducing fertility rates in Africa, models which describe the fertility patterns have not been adequately described. This article presents a flexible parametric model that can adequately capture the varying patterns of the age-specific fertility curves of African countries. The model has parameters that are interpretable in terms of demographic indices. The performance of this model was compared with other commonly used models and Akaike's Information Criterion was used for selecting the model with best fit. The presented model was able to reproduce the empirical fertility data of 11 out of 15 countries better than the other models considered. (ProQuest: ... denotes formulae omitted.) 1. Introduction Parametric and non parametric models have been reported to have useful applications in demographic research. Apart from being useful when creating hypothetical rate schedules in forecasting and projection, they also serve to condense complex data into smaller indices (Schmertmann 2003; Peristera and Kostaki 2007). Several models have; therefore, been proposed to model fertility as the major determinant (of the three demographic variables namely, fertility, mortality, and migration,); of the size and structure of any population. These models have been commonly created for the developed countries of the world and usually fit excellently the population they are intended to model (Hoem et al. 1981). It is pertinent, however, to mention that though there are many fertility models in the literature, few have been specifically generated to describe age-specific fertility patterns in Africa; despite the fact that most governments of the sub Saharan African countries are targeting lower total fertility rates to meet the Millennium Development Goals (MDGs) (United Nations 2000). To make reaching these targets possible, a better understanding of the current pattern of age specific fertility rate (ASFR) of African countries is required. Mathematical models, when well constructed, can aid in this understanding as they provide better insight into some characteristics of the distributional pattern of fertility in Africa. The goal of any modeling exercise is to extract as much information as possible from available data and to provide an accurate representation of both the known and unknown aspects of the phenomenon being studied (Salomon and Murray 2001). Modeling fertility in Africa has also become necessary to enable a meaningful comparison of fertility across the countries in the region in the face of the current fertility transition. Already, fertility can be compared using a wide variety of existing conventional measures, summary indices, or averages that are commonly reported for fertility data. These include total fertility rate (TFR), general fertility rate, and the crude birth rate. Few comparisons, however, are made based upon the detailed distribution of the age-specific fertility curve. Not all information in the curve can be conveyed by these summary indices. There is still much to be described in terms of the variance, skew, kurtosis, and symmetry of fertility distributions for individual countries on the continent. In this article, we propose a mathematical model for ASFR using the complementary error function (defined below). The proposed model is a flexible one that can capture various shapes of ASFR. It also provides a mathematical description of some fertility indices through its interpretable parameters. The efficacy of the model was determined by comparing its performance with other fertility models. The age pattern of fertility in Africa is described in the next section. In Section 3, we provide a brief review of some existing models for fertility patterns and then propose our model in Section 4. …