Global smartphone penetration has led to unprecedented addictive behaviors. To develop a smartphone use/non-use pattern by mobile application (App) in order to identify problematic smartphone use, a total of 79 college students were monitored by the App for 1 month. The App-generated parameters included the daily use/non-use frequency, the total duration and the daily median of the duration per epoch. We introduced two other parameters, the root mean square of the successive differences (RMSSD) and the Similarity Index, in order to explore the similarity in use and non-use between participants. The non-use frequency, non-use duration and non-use-median parameters were able to significantly predict problematic smartphone use. A lower value for the RMSSD and Similarity Index, which represent a higher use/non-use similarity, were also associated with the problematic smartphone use. The use/non-use similarity is able to predict problematic smartphone use and reach beyond just determining whether a person shows excessive use.