Simple, inexpensive and accurate analytical methods are in high demand. Dispersive solid-phase microextraction (DSPME) was used in combination with smartphone digital image colorimetry (SDIC) to determine boron in nuts as an approach replacing existing costly alternatives. A colorimetric box was designed to capture images of standards and sample solutions. ImageJ software was used to link pixel intensity to the analyte concentration. Under optimum extraction and detection conditions, linear calibration graphs were obtained with coefficients of determination (R2) above 0.9955. Percentage relative standard deviations (%RSD) were below 6.8 %. The limits of detection (LOD) ranged between 0.07 and 0.11 μg mL−1 (1.8 to 2.8 μg g−1), which were sufficient for detection of boron in nut samples (i.e., almond, ivory, peanut and walnut), with percentage relative recoveries (%RR) between 92.0 and 106.0 %.