San Francisco, CA tide chart

San Francisco's tide gauge stands at the Golden Gate, on the Presidio shore, where the whole of San Francisco Bay fills and empties through one strait. The tide there is mixed, with a mean range of about 4.1 feet and a lower low that averages about 1.1 feet below an average low.

Moving water sets the hour here. The NOAA gauge 5.4 km from the middle of San Francisco publishes 37 harmonic constants. Mean range is 4.1 ft, and the form factor — diurnal constituents over semidiurnal — is 0.84, which is mixed, mainly semidiurnal water.

Tide times and heights in San Francisco

The curve answers a question the table cannot. A table gives the minute of high water; the curve gives the hour you can actually get out — the stretch between one turn and the next that no row describes, and on a mixed, mainly semidiurnal coast that is most of the fishing day.

What the tide does at San Francisco

Along the city's bay shore the same tide arrives later the farther you go from the strait. NOAA's predictions for Pier 41 at North Point run about twelve minutes behind the Golden Gate, Pier 22½ at Rincon Point about twenty at high water and twenty-four at low, and Hunters Point about thirty and thirty-eight.

Ocean Beach runs the other way. NOAA's predictions for the outer coast there put high water about forty-nine minutes ahead of the Golden Gate and low water about thirty-five, so for the surf side of the city this table runs late by half an hour or more. Across the bay, Alameda runs about half an hour behind at high water and forty minutes at low.

What the heights are measured from

Every height here is feet above mean lower low water. That reference is not the sea bed and not the water under your keel — it is the average of the lower of each day's two low waters, taken over a nineteen-year epoch. A printed low water of 1.3 ft means the surface stands 1.3 feet above that line, not that there is 1.3 feet of water.

The useful consequence is that NOAA measures its charted depths from the same line. A sounding of four feet with a 1.3 ft tide is about 5.3 feet of water at that moment, and the same sounding at a −0.4 ft tide is about 3.6. Mixing the two references is the single most common way to read a tide table wrong.

Why the two high waters differ at San Francisco

The two high waters of a day are not twins at San Francisco. On average the higher one stands 0.61 ft above the mean high water, and the gap follows the Moon's declination rather than its phase: it is widest when the Moon stands far north or south of the equator and nearly closes when the Moon crosses it, about every two weeks.

The moon sets the rhythm and the sun decides how hard it lands. When the two pull in line the ranges stack; when they pull across each other the range collapses. At San Francisco that puts the spring range near 4.7 ft against a neap range near 2.9 ft — a factor of 1.62 between the strongest and the weakest tides of the fortnight.

Both figures come from the station's own constants rather than from the fortnight on this page, which is why they do not move when the table is rebuilt. They are also close to the station's published great diurnal range, and that agreement is a check on the arithmetic rather than a coincidence. The same calculation at Juneau, AK gives a mean range of 13.74 ft against 4.1 ft here — the same Moon and Sun acting on a different coast, with a different amount of water moving through it.

How close these numbers are to NOAA

These are not NOAA's published predictions. The station publishes 37 harmonic constants — amplitudes and phases of the periodic pulls that make up the tide here — and we synthesise the curve from them. Saying so is easy; the number that follows is the part that is worth reading.

Against NOAA's own predictions for August 12–19, 2026, every one of the 30 high and low waters matched, with an average difference in time of 0.55 minutes and a worst case of 1.44. The worst height difference was 0.6% of the day's range. That is the whole claim, and it is smaller than the width of the pencil line on a printed table.

Two limits belong with that number. It is a week of comparison, not a year, and it is this station: the same method run at a station with a big, awkward range does measurably worse, and our own worst case on the grid is off by minutes rather than seconds. Where that is true, the page says so, and the method behind every number on this site is written out rather than summarised.

When a tide table is not the right tool

A tide prediction is astronomy, and the water is weather as well. Wind and barometric pressure move the real level against the printed one: a hard onshore blow holds water in and makes a predicted low shallower and later, and a cold, dry high-pressure spell does the opposite. NOAA publishes observed levels alongside its predictions for exactly this reason. The numbers here are also the station's water, not the water two miles up a tidal creek, where the same tide arrives later and smaller. On rivers a spring freshet outweighs the tide outright.

Tide is the hour; it does not tell you which hour is worth taking. The reasoning behind the solunar windows — and what they are worth on a coast where moving water already sets the hour — is in best fishing times. Sun, moon and those windows for this same water, along with the monthly water temperature that decides what is feeding, are on the San Francisco, CA fishing forecast. The rest of the grid is catalogued on the fishing forecast hub. A printed page is also poor company in a creek at first light, and which fishing apps actually carry tide tables is a separate comparison.

The nearest stations this site carries are Alameda, CA (4.84 ft), Half Moon Bay, CA (3.85 ft), Monterey, CA (3.54 ft), Crescent City, CA (4.99 ft) and Honolulu, HI (1.28 ft), and the full list with the station behind each one is on the tide charts hub.