I’m trying to work on shticks, but my blogging time has been getting filled up with halakhic questions. And yes, Torah takes priority over Purim shticks.

Today is the Fast of Esther, which is observed a few days earlier than usual because of Shabbat. Around fast days, Rabbis constantly get asked what time the fast ends. When I give my answer, I usually hear the follow-up question, “Why are you 30 minutes (or more) earlier than everyone else?”
In Ta’anit 12a R. Hisda says that any fast which doesn’t last until sunset is not considered a fast. Logically then, a fast that does end at sunset is considered a fast, and sunset is therefore the minimal time for ending fast.

Rosh (Ta’anit 1:12) explains that sunset here doesn’t mean the beginning of sunset, but rather the end of sunset – called tzeit ha-kochavim the halakhic definition of nightfall. However, Rosh doesn’t provide any reason for this stringency.

Tosafot (Avoda Zara 34a s.v. Mit’anin L’shaot, Zevachim 56a s.v. Minayin L’dam, Menachot 20b s.v. Nifsal B’Shekiat) admit that the gemara in Ta’anit does mean the beginning of sunset, since elsewhere “sunset” means the beginning of sunset, but the practice developed to be stricter and to wait until nightfall. Sefer Kolbo (61) cites the Rif¹ as saying that many people end their fast immediately at sunset, following the gemara in Ta’anit. Despite this, Sefer Kolbo follows the custom of the Tosafot on the grounds that we are not experts in determining the conclusion of sunset, and is therefore strict to wait until nightfall.² Finally, Tur (Orach Hayim 562) and Beit Yosef (Orach Hayim 562:6) both define this instance of “sunset” as “nightfall”

The Vilna Gaon (OH 562:1) references Pesahim 54b where there is a debate as to whether fast of the 9th of Av ends at sunset, implying that it is certainly permitted on every other fast day. I have been told from several that the Vilna Gaon does rule the fast ends at sunset, but I have not seen the source inside.³
R. Ovadia Yosef acknowledges the inconsistent definitions of “sunset” as either referring to the beginning or end of sunset (Yehaveh Da’at 5:22) . He concludes that the generally accepted custom is to wait until nightfall, but those who wish to end earlier have on whom to rely (Yabia Omer 6 O.H. 31).

My sense is that the Bavli clearly defines a fast ending at sunset, not at nightfall. This is the most consistent reading of the Talmud; even Tosafot agree with this reading of the gemara in Ta’anit. Sometime between the Rif and Rosh, the custom changed and interpretations changed to match the custom.⁴ At any rate, the halakhic end for the fast is at sunset. The later stringency is not based on halakha, but on minhag (custom). This isn’t to say that minhagim could or should simply be ignored, but the talmud defines objective rules for how customs work within the halakhic system. Thus, the laws of fasting post-sunset are not inherent to the laws of fasting, but the laws of observing customs. At some point I’ll write more about that in detail, but until then see my shiur on minhagim and of course Rambam’s Introduction.

To find the exact times in your area, see the OU’s or kashrut.com’s zemanim calculators.

1. Kolbo doesn’t cite a source for this Rif, and a Bar Ilan search didn’t find it.
2. Rabbeinu Yerucham (Toldot Adam 18:1:163c), Tashbetz Katan (1), Sefer Yirei’im (274), Maharif (6), and a whole slew of others follow the stringency of Tosafot and Rosh.
3. Neither Bar Ilan nor the UC Library have been helpful.
4. I have no time to do a full History of Halakha report on this, but if you want to, I’d love to see what you find.