This paper investigates fractional share trading. We develop a methodology for identifying fractional share trades in the Consolidated Transaction Reporting System. Our approach uses a latency-based digital footprint to estimate fractional share trades executed by Robinhood and Drivewealth, the two largest fractional share broker dealers. We find a surprising breadth to fractional share trading: high-priced stocks, meme stocks, IPOs, SPACs, and other popular retail stocks now exhibit considerable numbers of these tiny trades. We show that these tiny trades matter: fractional share trades are predictive of future liquidity and volatility, suggesting an information content to fractional share trades. Our results indicate that our measure of fractional share trading better captures this market information than do standard measures of retail trading. We also discuss how current data and reporting protocols preclude knowing the full extent of fractional share trading, inflate trading volume data, and provide at best censored samples of these off-exchange trades.