In the cryptographic field, it is a cryptosystem practice to use a secret key together with a cryptographic algorithm parameter. These cryptographic algorithms should produce pseudorandom data to understand better the cryptographic performance of such ciphers. To understand the performance of a cryptographic algorithm a statistical analysis should be performed. In this paper, a randomness analysis of the proposed cryptographic algorithm using combined multilevel substitution and encryption techniques were executed. Statistical test results were presented using NIST STS to measure the randomness behavior of the keystream output. The statistical analysis utilized the non-parameterized test selection, such as cumulative sums, runs test, longest runs of ones, and frequency. The performed experiment used 100 samples with 100 bits in length. The algorithm achieved an average success rate of 99.50% in four statistical tests used. Therefore, it is assumed that the proposed cryptographic algorithm is effective in producing a random ciphertext sequence. For future works, it is then suggested to implement it in software application to boost electronic medical records protection.