The correlation between Standard and Cone Penetration Tests (SPT and CPT) as two of the most used in-situ geotechnical tests is of practical interest in engineering designs. In this paper, new SPT-CPT correlations for southwest of Sweden are proposed and developed using an artificial neural networks (ANNs) approach. The influences of soil type, depth, cone tip resistance, sleeve friction, friction ratio and porewater pressure on obtained correlations has been taken into account in optimized ANN models to represent more comprehensive and accurate correlation functions. Moreover, the effect of particle mean grain size and fine content were investigated and discussed using graph analyses. The validation of ANN based correlations were tested using several statistical criteria and then compared to existing correlations in literature to quantify the uncertainty of the correlations. Using the sensitivity analyses, the most and least effective factors on CPT-SPT predictions were recognized and discussed. The results indicate the ability of ANN as an attractive alternative method regarding to conventional statistical analyses to develop CPT-SPT relations.