This study aimed to evaluate psychometric properties of Functional Assessment of Cancer Therapy Cognitive Version 3 (FACT-Cog3). Scoring was compared when including then excluding four multitasking items that the scale authors added after validating the scale. This was intended to improve guidance about use of FACT-Cog3 in both clinical and non-clinical samples. Data from previous studies in people with and without cancer were supplemented with a new sample, for a total of 205 participants with and 110 participants without a history of cancer. Factors, reliability, and validity were examined in conjunction with other relevant measures. Exploratory factor analysis results supported a four-factor solution (i.e., perceived cognitive impairment [PCI], comments from others, perceived cognitive ability [PCA], and quality of life) for both versions of the scale in a clinical sample. In the non-clinical sample, factor analysis results suggested a four-factor solution for the 37-item scale but not the 33-item version, suggesting that the scale may perform differently in a non-clinical sample. High reliability and validity were found for both samples in the 33-item and 37-item versions of the scale. To gain a comprehensive understanding of perceived cognitive functioning it is recommended that the 37-item FACT-Cog3 scoring should be used in research and clinical practice. Most importantly, regardless of which scoring is used, it is essential that users clearly disclose which scoring method for the FACT-Cog3 has been selected and it is recommended that subscales are labelled accordingly (e.g., PCI18, PCA20, PCA7, PCA9).